Upgrade spacy to 3.7.4. Dockerfile rewrites

.envrc

+use_nix
+export LANG=C.UTF-8

Dockerfile

@@ -8,11 +8,18 @@ WORKDIR /tmp
 
 # Install dependencies:
 COPY ./requirements.txt .
-RUN pip install -r requirements.txt
+RUN pip install -r requirements.txt -f https://download.pytorch.org/whl/torch_stable.html
 
+# Build args
+# https://docs.docker.com/build/guide/build-args/
+#ARG PORT 8001
 # https://spacy.io/usage/models/
+ARG MODEL en_core_web_trf
+
+RUN python -m spacy download ${MODEL}
+
 #RUN python -m spacy download en_core_web_trf
-RUN python -m spacy download fr_dep_news_trf    # 8001
+#RUN python -m spacy download fr_dep_news_trf    # 8001
 #RUN python -m spacy download de_dep_news_trf   # 8002
 #RUN python -m spacy download uk_core_news_trf   # 8003
 #RUN python -m spacy download ru_core_news_trf   # 8004
@@ -30,4 +37,5 @@ WORKDIR /code
 
 COPY src/main.py .
 
-CMD ["uvicorn", "main:app", "--host", "0.0.0.0", "--port", "8001"]
+ENV SPACY_MODEL=${MODEL}
+CMD uvicorn main:app --host 0.0.0.0 --port 8000

default.nix

+with (import <nixpkgs> {});
+stdenv.mkDerivation rec {
+  name = "env";
+  env = buildEnv {
+    name = name;
+    paths = buildInputs;
+  };
+  buildInputs = [
+    docker-compose
+  ];
+}

docker-compose.yaml

 services:
-  spacy-server-de:
-    build: ./
-    ports:
-      - 8002:8002
-
   spacy-server:
-    build: ./
+    build:
+      context: ./
+      args:
+        MODEL: en_core_web_trf
     ports:
-      - 8001:8001
+      - 8000:8000
+      
+  spacy-server-fr:
+    build:
+      context: ./
+      args:
+        MODEL: fr_dep_news_trf
+    ports:
+      - 8001:8000
+
+  # spacy-server-de:
+  #   build:
+  #     context: ./
+  #     args:
+  #       MODEL: de_dep_news_trf
+  #   ports:
+  #     - 8002:8000
+
+  # spacy-server-uk:
+  #   build:
+  #     context: ./
+  #     args:
+  #       MODEL: uk_core_news_trf
+  #   ports:
+  #     - 8003:8000
+
+  # spacy-server-ru:
+  #   build:
+  #     context: ./
+  #     args:
+  #       MODEL: ru_core_news_trf
+  #   ports:
+  #     - 8004:8000
+
+  # spacy-server-pl:
+  #   build:
+  #     context: ./
+  #     args:
+  #       PORT: 8004
+  #       MODEL: pl_core_news_trf
+  #   ports:
+  #     - 8004:8000
+
+  # spacy-server-es:
+  #   build:
+  #     context: ./
+  #     args:
+  #       PORT: 8005
+  #       MODEL: es_core_news_trf
+  #   ports:
+  #     - 8005:8000
+
+  # spacy-server-pt:
+  #   build:
+  #     context: ./
+  #     args:
+  #       PORT: 8006
+  #       MODEL: es_core_news_trf
+  #   ports:
+  #     - 8006:8000
+
+  # spacy-server-it:
+  #   build:
+  #     context: ./
+  #     args:
+  #       PORT: 8007
+  #       MODEL: it_core_news_trf
+  #   ports:
+  #     - 8007:8000
+
+  # spacy-server-zh:
+  #   build:
+  #     context: ./
+  #     args:
+  #       PORT: 8008
+  #       MODEL: it_core_news_trf
+  #   ports:
+  #     - 8008:8000
+
+  # spacy-server-el:
+  #   build:
+  #     context: ./
+  #     args:
+  #       PORT: 8009
+  #       MODEL: el_core_news_trf
+  #   ports:
+  #     - 8009:8000
+
+  # spacy-server-xx:
+  #   build:
+  #     context: ./
+  #     args:
+  #       PORT: 8010
+  #       MODEL: xx_ent_wiki_trf
+  #   ports:
+  #     - 8010:8000
 

requirements.txt

@@ -4,11 +4,16 @@
 
 spacy-transformers
 setuptools
-numpy==1.24.2
-sense2vec==2.0.0
-ujson==4.2.0
-fastapi==0.55.0
-uvicorn==0.15.0
+fastapi==0.110.0
+numpy==1.26.4
 pytest==6.2.5
-requests==2.26.0
-spacy==3.1.3
+sense2vec==2.0.2
+six==1.16.0
+spacy==3.7.4
+uvicorn==0.28.0
+
+# If you need GPU support, just change the following lines
+# https://stackoverflow.com/questions/68845314/install-pytorch-cpu-only-in-dockerfile
+# https://download.pytorch.org/whl/torch_stable.html
+# torch==2.2.1
+torch==2.2.1+cpu

src/main.py

@@ -6,20 +6,20 @@ from typing import List
 import spacy
 from dataclasses import dataclass
 from fastapi import FastAPI, HTTPException
-from pydantic import BaseModel, root_validator
+from pydantic import BaseModel, model_validator
 from sense2vec import Sense2VecComponent
 from starlette.responses import Response
 from starlette.status import HTTP_204_NO_CONTENT
 
 # download model before
-import fr_core_news_md
+#import fr_core_news_md
 
 app: FastAPI = FastAPI()
 
-#model: str = os.getenv('SPACY_MODEL')
+model: str = os.getenv('SPACY_MODEL') or 'en_core_web_trf'
 #pipeline_error: str = f"The model ({model}) doesn't support " + '{}.'
-#nlp: spacy = spacy.load(model)
-nlp: spacy = fr_core_news_md.load()
+nlp: spacy = spacy.load(model)
+#nlp: spacy = fr_core_news_md.load()
 
 if os.getenv('SENSE2VEC') == '1':
     s2v = nlp.add_pipe("sense2vec")
@@ -131,7 +131,7 @@ class PhraseInSentence(BaseModel):
     sentence: str
     phrase: str
 
-    @root_validator
+    @model_validator(mode='after')
     def phrase_must_be_in_sentence(cls, values):
         if values.get('phrase') not in values.get('sentence'):
             raise ValueError('phrase must be in sentence')
@@ -195,7 +195,7 @@ class Token:
     tag: str
     dep: str
     lang: str
-    prob: int
+    prob: float
     char_offset: int